WO1996032579A1 - Process for finding the mass of air entering the cylinders of an internal combustion engine with the aid of a model - Google Patents

Abstract

Calculating the mass of air actually flowing into the cylinder with the aid of an inlet pipe filling model which, from the factor angle of throttle valve aperture, environmental pressure and parameters representing the valve control, provides a load value on the basis of which the injection time is determined. In addition, this load value is used in predictions for estimating the load value at a time at least one sampling step later than the actual calculation of the injection time.

Description

description

Method for model-based determination of the air mass flowing into the cylinders of an internal combustion engine

The invention relates to a method for model-based determination of the air mass flowing into the cylinders of an internal combustion engine according to the preamble of claim 1.

Engine control systems for internal combustion engines using

Working fuel injection, require the air mass mzyi sucked in by the engine as a measure of the engine load. This parameter forms the basis for realizing a required air-fuel ratio. Growing demands on engine control systems, such as the reduction in pollutant emissions from motor vehicles, mean that the load size for stationary and unsteady-state processes must be determined with low permissible errors. In addition to the mentioned operational cases, the precise load detection during the warm-up phase of the internal combustion engine offers considerable potential for reducing pollutants.

In air mass-guided engine control systems, the signal of the air mass meter, which is used as the load signal of the internal combustion engine, and which is arranged upstream of the intake manifold, does not represent a measure of the actual filling of the cylinders in stationary operation because the volume of the intake manifold acts downstream of the throttle valve as an air reservoir that has to be filled and emptied. The decisive air mass for the injection time calculation is that air mass which flows out of the intake manifold and into the respective cylinder.

In intake manifold pressure-guided engine control systems, the output signal of the pressure sensor reflects the actual pressure conditions in the intake manifold, but the measured variables are available Due to the necessary averaging of the measured variable, inter alia, it is available relatively late.

With the introduction of variable intake systems and variable valve controls, a very large number of influencing variables which influence the corresponding model parameters arise for empirically obtained models for obtaining the load size from measurement signals.

Model-based calculation methods based on physical approaches represent a good starting point for the precise determination of the air mass mzyi.

DE 39 19 488 C2 discloses a device for regulating and predicting the intake air quantity of an internal combustion engine guided by intake manifold pressure, in which the degree of throttle valve opening and the engine speed are used as the basis for calculating the current value of the air drawn into the combustion chamber of the engine . This calculated, current amount of intake air is then used as the basis for calculating the predetermined value for the amount of intake air to be drawn into the combustion chamber of the engine at a specific time from the point at which the calculation was carried out , used. The pressure signal which is measured downstream of the throttle valve is corrected with the aid of theoretical relationships, so that an improvement in the determination of the intake air mass is achieved and a more precise calculation of the injection time is possible.

In transient operation of the internal combustion engine, however, it is desirable to carry out the determination of the air mass flowing into the cylinders even more precisely.

The invention is based on the object of specifying a method with which the air mass actually flowing into the cylinder of the internal combustion engine can be determined with high accuracy. In addition, system-related dead time Ten, which can occur due to the fuel storage and the computing time when calculating the injection time, are compensated.

This object is achieved according to the features of patent claim 1.

Advantageous further developments can be found in the subclaims.

Starting from a known approach, a model description results which is based on a nonlinear differential equation. An approximation of this nonlinear equation is presented below. As a result of this approximation, the system behavior can be described by means of a bilinear equation which allows the relationship in the engine control unit of the motor vehicle to be quickly resolved under real-time conditions. The selected model approach includes the modeling of variable suction systems and systems with variable valve controls. The by this arrangement and by dynamic reloading, i.e. Effects caused by reflections of pressure waves in the intake manifold can only be very well taken into account solely by the choice of stationary parameters of the model. On the one hand, all model parameters can be interpreted physically and, on the other hand, they can only be obtained from stationary measurements.

Most algorithms for the time-discrete solution of the differential equation, which describes the behavior of the model used here, require a very small computing step size, in particular if the pressure drop across the throttle valve is low, ie at full load, in order to work numerically stable. The consequence would be an unacceptable computing effort when determining the load size. Since load detection systems mostly work segment-synchronously, ie for 4-cylinder engines a measured value is sampled every 180 ° KW, the model equation must also be segment be solved synchronously. In the following, an absolutely stable difference scheme is used to solve differential equations, which guarantees numerical stability with any step size.

The model-based calculation method according to the invention also offers the possibility of predicting the load signal by a selectable number of sampling steps, i.e. a prediction of the load signal with a variable prediction horizon. If the prediction time proportional to the prediction horizon at constant speed does not become too long, a predicted load signal of high accuracy is obtained.

Such a prediction is necessary because a dead time arises between the acquisition of relevant measured values and the calculation of the load size. Furthermore, for reasons of mixture preparation, the fuel mass must be metered as precisely as possible via the injection valves before the actual start of the intake phase of the respective cylinder, which is in the desired ratio to the air mass m _cyl in the course of the upcoming intake phase. A variable prediction horizon improves the quality of the fuel metering in transient engine operation. Since the segment time decreases with increasing speed, the injection process must start a larger number of segments sooner than at a lower speed

Case is. In order to be able to determine the fuel mass to be metered as precisely as possible, the prediction of the load size by the number of segments by which the fuel is stored is necessary in order to maintain the required air-fuel ratio in this case too. The prediction of the load size thus contributes from a substantial improvement in compliance with the required fuel-air ratio in transient engine operation. This system for model-based load detection is in the known engine control systems, that is to say in the case of air mass-guided or intake-pipe pressure-guided engine control systems, a correction algorithm in the form of a model control loop is described below formulated, which allows permanent accuracy improvement, ie a model comparison in stationary and transient operation, when inaccuracies of model parameters occur.

An embodiment of the method according to the invention is described with reference to the following schematic drawings. Show:

1 shows a schematic diagram of the intake system of an Otto engine, including the corresponding model and measurement variables, FIG. 2 shows the flow function and the associated polygonal approximation, Intake manifold pressure-controlled engine control systems.

Λ The model-based calculation of the load variable mzyi is based on the basic arrangement shown in FIG. 1. For reasons of clarity, only one cylinder of the internal combustion engine is shown. The reference numeral 10 denotes an intake manifold of an internal combustion engine, in which a throttle valve 11 is arranged. The throttle valve 11 is connected to a throttle valve position sensor 14 which determines the degree of opening of the throttle valve. An air mass meter 12 is arranged upstream of the throttle valve 11 in an air mass-guided engine control system, while an intake manifold pressure sensor 13 is arranged in the intake manifold in an intake manifold pressure-guided engine control system. Depending on the type of load detection, only one of the two components 12, 13 is therefore present. The outputs of the air mass meter 12, the throttle valve position sensor 14 and the intake pipe pressure sensor 13, which is available as an alternative to the air mass meter 12, are provided with inputs of an electronic control device, which is not shown and is known per se. 96/32579 PCI7DE96 / 00615

tion of the internal combustion engine connected. In addition, an inlet valve 15, an outlet valve 16 and a piston 18 movable in a cylinder 17 are shown schematically in FIG.

In addition, selected sizes or parameters of the suction system are shown in FIG. The roof symbol "Λ" above a size means that it is a model size, while sizes without a roof symbol "Λ" represent measured values. In particular:

Pu ambient pressure, Pg intake manifold pressure, Tg temperature of the air in the intake manifold, Vg the volume of the intake manifold.

Sizes with a dot symbol indicate the first time

Derivation of the corresponding quantities, YΠDK is thus the air mass flow at the throttle valve and mzyi is the air mass flow that actually flows into the cylinder of the internal combustion engine.

The basic task in the model-based calculation of the engine load state now consists in solving the differential equation for the intake manifold pressure

die sich unter der Voraussetzung konstanter Temperatur der Luft im Saugrohr Tg aus der Zustandsgieichung idealer Gase herleiten läßt.

which can be derived from the condition of ideal gases under the condition of constant air temperature in the intake manifold Tg.

The general gas constant is denoted by RL.

Λ

The load size mzyi is determined by integration from the cylinder

Λ mass flow mzyi determined. The relationships described by (2.1) are common to multi-cylinder internal combustion engines Vibrating tube (switching suction tube) and / or resonance suction systems applicable without structural changes.

For systems with multi-point injections, in which the fuel is metered through several injection valves, equation (2.1) gives the conditions more accurately than with single-point injections, i.e. in the case of injections in which the fuel is metered by means of a single fuel injection valve. With the first-mentioned type of fuel metering, almost the entire intake system is filled with air. There is only a fuel-air mixture in a small area in front of the inlet valves. In contrast, in single-point injection systems, the entire intake manifold from the throttle valve to the intake valve is filled with a fuel-air mixture, since the injection valve is arranged in front of the throttle valve. In this case, the assumption of an ideal gas represents a closer approximation than is the case with multi-point injection. With single-point injection, the fuel is metered accordingly

Λ Λ mDK, corresponding to mzyi for multi-point injection.

Λ

The calculation of the mass flows IΠDK and

Λ mzyi described in more detail.

The model size of the air mass flow at the throttle valve

Λ mDK is described by the flow equation of ideal gases through throttling points. Flow losses occurring at the throttle point are reduced by the reduced flow cross

Λ Λ section ARED considered. The air mass flow mDK is therefore determined by the relationship

mit

With

for supercritical pressure conditions

respectively.

ψ = const. for critical pressure conditions (2.2)

certainly .

Λ mDK: model size of the air mass flow at the throttle valve

Λ

A RED reduced flow cross section

K: Adiabatic exponent

R _L : general gas constant

T _S = temperature of the air in the intake manifold

Λ

Pu: Mode11 size of the ambient pressure

Λ

Ps: model size of the intake manifold pressure

Ψ: flow function.

At the throttle point, i.e. occurring on the throttle valve

Λ

Flow losses are taken into account through the appropriate choice of ARED. With known pressures in front of and behind the throttle point and known mass flow through the throttle point, an assignment between the throttle valve angle determined by the throttle valve position sensor 14 and the corresponding reduced cross section can be made from stationary measurements

Λ A RED can be specified.

If the air mass flow mDK at the throttle valve is described by the relationship (2.2), a more complex one arises Algorithm for the numerically correct solution of the differential equation (2.1). To reduce the computational effort, the flow function ψ is approximated by a polygon.

FIG. 2 shows the course of the flow function ψ and the approximation principle applied to it. Within a section i (i = l ... k) the flow function ψ is represented by a straight line. With a reasonable number of straight line sections, a good approximation can be achieved. With such an approach, equation (2.2) can be used to calculate the mass flow at the throttle valve

Λ mDK through the relationship

mDK APPROX

can be approximated for i = (l ... k).

In this form, mj_ describes the slope and nj_ the absolute term of the respective line segment. The values for the slope and for the absolute member are shown in tables as a function of the ratio of intake manifold pressure to ambient pressure

The abscissa of Figure 2 is the pressure ratio

Λ p

- Λ and on the ordinate the function value (0 - 0.3) of the

Pυ

Flow function ψ plotted.

Λ K

P 2 "1κ-l

For pressure conditions —— <ψ = constant, i.e.

_P + u that the flow at the throttle point is only dependent on the cross-section and no longer on the pressure ratio sen. The air mass flowing into the respective cylinders of the internal combustion engine is difficult to determine analytically, since it is strongly dependent on the gas exchange. The filling of the cylinders is largely determined by the intake manifold pressure, the speed and the valve timing.

To calculate the mass flow in the respective

Because of the cylinder mzyi, it is necessary on the one hand to describe the conditions in the intake tract of the internal combustion engine by means of partial differential equations and on the other hand to calculate the mass flow at the inlet valve according to the flow equation as a necessary boundary condition. It is only this complicated approach that allows dynamic reloading effects to be taken into account, which are significantly influenced by the speed, the intake manifold geometry, the number of cylinders and the valve control times.

Since a calculation based on the above-mentioned approach cannot be implemented in the electronic control device of the internal combustion engine, a possible approximation is based on a

Achen multiple relationship between intake manifold pressure P _s and cylinder as-

Strom senstrom mzyi. For a wide range of useful valve timing, a linear approach to this can be taken as a good approximation

be assumed.

The slope γ _] and the absolute member γ _{0 of} the relationship (2.4) are functions of the speed, the intake manifold geometry, the number of cylinders, the valve timing and the temperature of the air in the intake manifold Tg, taking into account all essential influencing factors. The dependence of the values on γj and γ ₀ from the influencing variables speed, intake manifold geometry, The number of linders and the valve timing and valve lift curves can be determined using stationary measurements. The influence of vibrating tube and / or resonance suction systems on the air mass sucked in by the internal combustion engine is also well reproduced via this value determination. The values of> and γ ₀ are stored in characteristic diagrams of the electronic engine control device.

The intake manifold pressure Pg is selected as the determining variable for determining the engine load. With the help of the model differential equation, this quantity should be as precise and fast as possible

Λ can be estimated. The estimation of Ps requires the solution of equation (2.1).

With the simplifications introduced using formulas (2.2) and (2.3), (2.1) can be determined by the relationship

für i = (1...k) (2.5)

for i = (1 ... k) (2.5)

be approximated. If, in accordance with the requirements for deriving equation (2.1), the temperature of the air in the intake manifold Tg is considered to be a slowly changing measurement

Λ large and ARED as the input variable, the non-linear form of the differential equation (2.1) can be approximated by the biliary equation (2.5).

To solve equation (2.5), this relationship is converted into a suitable difference equation.

The following basic requirements for the solution properties of the difference equation to be formed can be formulated as a criterion for selecting the suitable difference scheme: 1. The difference scheme must be conservative even under extreme dynamic requirements, ie the solution of the difference equation must correspond to the solution of the differential equation,

2. The numerical stability must be guaranteed in the entire working range of the intake manifold pressure at sampling times which correspond to the maximum possible segment times.

Claim 1 can be met by an implicit calculation algorithm. Due to the approximation of the nonlinear differential equation (2.1) by means of a bilinear equation, the resulting implicit solution scheme can be solved without using iterative methods, since the difference equation can be converted into an explicit form.

Due to the conditioning of the differential equation (2.1) and its approximation (2.5), the second requirement can only be met by a calculation rule for the formation of the difference equation, which works absolutely stable. These processes are also referred to as A-stable processes. Characteristic of this A-stability is the property of the algorithm, with a stable output problem for any values of the sampling time, i.e. Segment time T ^ to be numerically stable. A possible calculation rule for the numerical solution of differential equations that meets both requirements is the trapezoidal rule.

The difference equation resulting from the application of the trapezoid rule is in the present case

^p _s w = ^pΛ _s [tf-ι] + γ- Ps [ ^ΛT - ^I ] ^{+ P} ; w

for N = (1.. cP) (2. 6)

Are defined If this rule is applied to (2. 5), the relationship results

for N = (1 ... 0P) and i = (l ... k) (2.7)

Λ for calculating the intake manifold pressure P _s [N] as a measure of the engine load.

[N] means the current segment or the current arithmetic step, [N + l] the next following segment or the next arithmetic step.

The calculation of the current and predicted load signal is described below.

Λ

From the calculated intake manifold pressure Ps, the air mass

Λ flow mzyi that flows into the cylinders can be determined using the relationship (2.4). If a simple integration algorithm is used, the relationship is obtained for the air mass sucked in by the internal combustion engine during an intake stroke

for N = (1.. Xfi) (2. 8) It is assumed that the initial value of the load size is zero. For segment-synchronous load detection, the segment time decreases with increasing speed, while the number of segments by which the fuel is stored must increase. For this reason, it is necessary to design the prediction of the load signal for a variable prediction horizon H, ie for a specific number H of segments, which is primarily speed-dependent. If this variable prediction horizon H is taken into account, equation (2.8) can be in the form

für N = ( l . . CP ) ( 2 . 9 )

for N = (1st CP) (2nd 9)

to be written.

For further considerations, it is assumed that the segment time T ^ and the parameters γ, and γ _{0 of} the loading

Λ drawing (2.4) used to determine the mass flow mz _y ι

Λ the intake manifold pressure P _{s are} required, do not change over the prediction time. Under this condition, the prediction of a value

Λ for mz _y ι [N + H by predicting the corresponding pressure

Λ reached Ps [N + H]. As a result, the equation (2.9) takes the form

für N = (l... P) (2.10)mz _≠ [N] m- + 2 -y ₀

for N = (l ... P) (2.10)

on.

Since in the described method the time change of the

Λ

Intake manifold pressure Ps is in analytical form, is in

Λ are followed by the prediction of the pressure value Ps [N + H] by H- multiple application of the trapezoidal rule achieved. In this case the relationship is obtained

Ps [N + H] = Ps [N] + ^ - - H - _s [N - \] + P _s [N]

for N = (! .. <») (2.11)

If the pressure Ps [N + H- l] is determined in an analogous manner, the equation can be used for the predicted load signal

mzy, [N + H] = T _A - γ, _s [ _N ] + (HQ.5) - Ps [N - l] + Ps W ^• + Yo

for N = (1 .. «00) (2.12)

can be specified.

If values of the order of 1 to 3 segments are selected for the prediction horizon H, a well-predicted load signal can be obtained with the formula (2.12).

The principle of model matching for air mass and intake manifold pressure-guided engine control systems is explained below.

Due to the use of engines with variable valve control and / or variable intake manifold geometry

Manufacturing tolerances and signs of aging, as well as temperature influences, the values of y <and γ _{0 are subject} to a certain degree of uncertainty. As described above, the parameters of the equation for determining the mass flow in the cylinders are functions of various influencing variables, of which only the most important ones can be recorded.

When calculating the mass flow at the throttle valve, measurement errors affect the detection of the throttle valve angle and approximation errors in the polygon approximation tion of the flow function ψ on the model sizes. In the case of small throttle valve angles in particular, the system sensitivity to the first-mentioned errors is particularly high. It follows from this that small changes in the throttle valve position have a serious influence on the mass flow or intake manifold pressure. In order to reduce the effect of these influences, a method is proposed below which makes it possible to correct certain variables which have an influence on the model calculation in such a way that an accuracy-improving model adaptation for stationary and unsteady engine operation is carried out can.

Essential parameters of the model for determining the load size of the internal combustion engine are adjusted by correcting the throttle valve angle measured from the

Λ agreed reduced cross-section ARED through the correction

AA large AA RED ■

The input variable for the corrected intake manifold pressure calculation

Λ ARED becomes through the relationship

A REDKORR = ARED + A A RED (3.11)

described.

Then in equation (2.2) and the following formulas

Λ Λ

ARED replaced by AREDKORR. To improve the follow-up behavior of the control loop, the reduced throttle valve cross-section derived from the measured value of the throttle valve angle is

Λ ARED included in the model calculation. The corrective

Door size AARED is formed by implementing a model control loop. For air mass-guided engine control systems, the air mass flow mDK_LMM measured by means of the air mass meter on the throttle valve is the reference variable of this control loop, while the intake manifold pressure P _s measured is used as the reference variable for intake manifold pressure-guided systems. About a fol-

Λ control, the value of AARED is determined in such a way that the control deviation between the reference variable and the corresponding control variable is minimized.

In order to achieve accuracy improvements with the aforementioned method even in dynamic operation, the measured value acquisition of the reference variable must be reproduced as precisely as possible. In most cases, the dynamic behavior of the sensor, i.e. either the air mass meter or the intake manifold pressure sensor and a subsequent averaging.

The dynamic behavior of the respective sensor can be modeled in a first approximation as a first-order system with possibly working point-dependent delay times T_. In the case of an air mass-guided system, one possible equation is to describe the sensor behavior

rriDκ_LMM [N] = e ^{τ> ■} TΠDK [N - rTlDK_LMM [N - 1] (3. 12]

A size that is essential in the chosen approach

Λ flow to the maximum possible mass flow mzyi is

Λ the ambient pressure Pυ. For this reason, a constant value of this size cannot be assumed, but an adjustment is made in the manner described below.

Λ

The value of the ambient pressure Pυ is changed when the amount of the correction variable A ARED reaches a certain threshold Λ or if the pressure ratio - Ps is greater than

Pυ is a selectable constant. This ensures that an ambient pressure adjustment can take place both in the partial and in the full-load range.

A model comparison for air mass-guided engine control systems is explained below. The model structure shown in FIG. 3 can be specified for this system.

The throttle position sensor 14 (Figure 1) provides a signal corresponding to the degree of opening of the throttle valve 11, e.g. a throttle opening angle. In a map of the electronic engine control device, values for the reduced cross section of the throttle valve opening are associated with various values of this throttle valve opening angle.

Λ flap ARED saved. This assignment is represented by the block "static model" in FIG. 3 and in FIG. 4. The “intake manifold model” subsystem in FIGS. 3 and 4 represents the behavior described by (2.7). The leading variable of this model control loop is the measured value of the air mass flow at the throttle averaged over a segment

flap mDK_LMM • If a PI controller is used as a controller in this model control loop, the remaining control deviation is zero, i.e. Model size and measured quantity of the air mass flow at the throttle valve are identical.

The pulsation phenomena of the air mass flow at the throttle valve, which can be observed especially in 4-cylinder engines, lead to considerable positive measurement errors in the case of air mass meters which form the amount, and thus to a command variable which is severely faulty. By switching off the

Controller, ie a reduction in the controller parameters can be switched to controlled model-based operation. Areas in which the pulsations mentioned can thus be treated with the same method, taking into account dynamic relationships, as those rich, in which there is an almost undisturbed leader. In contrast to methods that only take relevant measured values into account at stationary operating points, the system described remains operational almost without restrictions. If the air mass signal or the signal from the throttle valve position sensor fails, the system presented is able to generate a corresponding replacement signal. If the command variable fails, the controlled operation must be implemented, while in the other case the regulated operation guarantees the hardly impaired functionality of the system.

The "intake manifold model" block represents the conditions as described using equation (2.7) and has therefore

Λ the model size P _s and the time

Λ Λ the derivative P _s and the size mDK. ■ After modeling the sensor transmission behavior, ie the transmission behavior of the air mass meter and the scanning, the model

Λ large mDκ_LMM subjected to averaging so that the

Λ the middle variable mDK_LMM and the average air mass flow mDK_LMM measured by the air mass meter can be fed to a comparator. The difference between the two signals causes

Λ AARED change in the reduced flow cross-section

Λ

ARED, SO that a model comparison can be made stationary and unsteady.

The model structure shown in FIG. 4 is given for intake manifold pressure-guided engine control systems, the same blocks as in FIG. 3 being given the same designations. As with the air mass-guided engine control system, the “intake manifold model” subsystem represents the behavior described by the differential equation (2.7). The reference variable of this model control loop is the measured value of the intake manifold pressure Ps_s averaged over a segment. If a PI controller is also used, as in FIG. 3, then in the stationary case 96/32579 PCIYDE96 / 00615

20 the measured value of the pressure in the intake manifold Ps_s with the model size

Λ

Ps_s identical. As described above, the present system also remains fully operational, since a corresponding substitute signal can be generated if the intake manifold pressure signal or the measured value for the throttle valve angle fails.

The model variables P _s , P _s obtained by the intake manifold model are fed to a "prediction" block. Since the models also calculate the pressure changes in the intake manifold, these pressure changes can be used to determine the future pressure curve in the intake manifold and thus the cylinder air mass for the next [N + 1] or for the next segments [ N + H] TO APPRECIATE. The size mz _y ι and the size mz _y ι [N + \] then serve for the exact calculation of the injection time during which fuel is injected.

Claims

claims 1. Method for determining the inflowing air mass into the cylinder or cylinders of an internal combustion engine with an intake system which has an intake manifold (10) and a throttle valve (11) arranged therein, and a throttle valve position which detects the degree of opening of the throttle valve (19) lungs sensor (14), - A sensor (12; 13) generating a load signal (mDK_LMM \ Ps_s) from the internal combustion engine - An electrical control device on the On the basis of the measured load signal (mDκ_LMM ', Ps_s) and the speed of the internal combustion engine, a basic injection time is calculated, that is to say that - the conditions in the intake system are simulated by means of an intake manifold filling model, the parameters representing the opening degree of the throttle valve (11), the ambient pressure (P _v ) and the valve position being used as input variables of the model, Λ a model size for the air mass flow ( ^* wz) A .-) at the throttle valve (11) is described using the flow equation of ideal gas through throttle points (equation 2.2), Λ - A model size for the air mass flow (mzyi) in or into the cylinder (17) as a linear function of the intake manifold Λ pressure (Ps) by a mass balance of the air mass flows Λ Λ (mDK, mzyi) is described (Equation 2.1) - These model variables are linked via a differential equation (equation 2.5), from which the intake manifold pressure (Ps) is calculated as a determining variable for determining the actual load of the internal combustion engine (equation 2.7) and - from the linear relationship (equation 2.4) between Λ calculated intake manifold pressure (Ps) and the model size for the Λ Lu mass flow (mz _y ι) into or into the cylinders (17) is obtained by integrating the air mass (mz _y ι) flowing into Λ or into the cylinders (17). 2. The method of claim 1, d a d u r c h g e k e n n z e i c h n e t that the load signal measured by the load sensor (12; 13) (mDK_LMM; Ps_s) for correction and thus for comparison of the mo- Λ dell variables (mz _y ι) are used in a closed control loop, the load signal (moκ_LMM; Ps_s) serving as the command variable of the control loop. 3. The method of claim 2, d a d u r c h g e k e n n z e i c h n e t that the adjustment is carried out in the stationary and / or transient operation of the internal combustion engine and taking into account the transmission behavior of the load sensor (12; 13). 4. The method of claim 2, d a d u r c h g e k e n n z e i c h n e t that each measured value of the throttle valve opening degree ei Λn value of a reduced cross section of the throttle valve (ARED) and the comparison of the model sizes by correcting the reduced cross-section (ARED) Λ by means of a correction variable (AARED) such that the control deviation between the reference variable and the corresponding model variable is minimized. 5. The method according to claim 4, so that the reduced cross-section (ARED) is determined from stationary measurements on the engine test bench and is stored in a map of a memory of the electrical control device. 6. The method according to claim 1, 96/32579 PCI7DE96 / 00615 23 characterized in that when the model size for the air mass flow (mDK) is represented on the throttle valve (11), a flow function (ψ) existing in the flow equation (equation 2.2) into individual sections (i = l ... k) is divided and these sections are approximated by straight line sections, the slope (m ^) and the absolute term (n- ^) of the respective straight line sections as a function of the ratio Λ Λ are determined by intake manifold pressure (Ps) and ambient pressure (Pυ) and are stored in a map. 7. The method according to claim 1, characterized in that the slope (γ,) and the absolute member (γ ₀ ) of the linear function for the model size for the air mass flow in or ind the cylinder (mz _y ι) depending on at least one of the parameters , Speed of the internal combustion engine, number of cylinders, intake manifold geometry, temperature of the air (Tg) in the intake manifold (10) and valve control characters. 8. The method of claim 7, d a d u r c h g e k e n n z e i c h n e t that the parameters are determined by stationary measurements on the engine test bench and are stored in maps. 9. The method according to claim 1, characterized in that the air mass flowing into the cylinder (mz _y ι) by the relationship

is calculated with T _{} \} : sampling time or segment time mz _y ι [N] model size of the air mass flow during the current sampling step or segment mz _y t [N - \] model size of the air mass flow during the previous sampling step or segment. 10. The method of claim 1, d a d u r c h g e k e n n z e i c h n e t that the air mass flowing into or into the cylinder (mzyi) for a certain prediction horizon (H), which lies in the future with respect to the current load detection at the sampling time [N], is estimated by estimating the corresponding pressure value according to the following relationship: mzyi [N + H]

With T ^: sampling time or segment time H: prediction horizon, number of sampling steps γ in the future,: slope of the linear equation γ ₀ : absolute term for determining mzι N: current sampling step 11. The method according to claim 10, so that the number (H) of segments for which the load signal is to be estimated for the future is defined as a function of the speed.